COVID-19 Statistics

Simple repository to explore and show data about the new pandemic virus Corona Virus.

Datasets

Running notebook

In case you are running the notebook, do not forget to give a:

$> pip install -r requirements.txt

Download Firefox webdriver from https://github.com/mozilla/geckodriver/releases/tag/v0.26.0 to save the heatmap at the end.

Statistics

1. World data


The bellow graphics show the evolution of the desease over time for some countries.

Logarithmic Scale

Bellow some graphics in log scale of Confirmed cases for the above countries.


Predicting the pandemic of Corona Virus is hard, bellow is a simple demonstration of curve fitting, using 2 types (exponential and sigmoid) for estimation.


Another problem is not knowing the actual mortality for the disease.

The mortality for Covid-19 are said to be something like 3.8%, but previous calculations (based on data from China) put the mortality to be somthing like 2%... if this is truth, and looking at the mortality for Brazil, US and Italy, we should be able to extrapolate and calculate the possible real number of people who has the disease (been asyntomatic or not).


2. Cases for US

Simple compilation of cases in United States.


3. Cases for Brazil

Simple compilation of cases in Brazil.

For the entire Brazil, as of today, we have the following numbers:


But, the story can't be told by the entire country, one must take into account, each state of the federation. Let's show data for each state in the federation.

Each state tells a different story, but what about the capitals for some of those states?

Bellow some possible projections for the next 10 days of infected people for each capital showed above.


Deaths in each capital are growing... let's visualize how deaths are spread across some cities.

Much has been talked about that people in Brazil are young, so there's little risk for the population... but if we take into account that Brazil population is one of the biggest in the world and calculating death or hospitalization based on data provided by Imperial College and China CDC, Brazil could have more than 50k deaths. This is a simplistic view... it should take into account comorbidities to calculate those numbers.


Given the above values (from Imperial College and CDC China) lets do a projection of the possible # of deaths in each Age group and given a possible interval of deaths that may occur.